AWS ECR Push Action

Action to push images to Amazon's Elastic Container Registry.

Usage

Make sure to checkout the repo so the workflow can see your Dockerfile.

on: [push]

jobs:
  ecr:
    runs-on: ubuntu-latest
    name: Docker build and push to ECR
    env:
      ## Repository must be in the format '<BU>/<SQUAD>/<PROJECT NAME | MY NAME>'
      ECR_REPOSITORY: 'cross/devtools/momo'
      ECR_IMAGE_VERSION: '0.2.2'
    steps:

      - name: Checkout
        uses: actions/checkout@v4

      # Exemple of build using docker
      - name: Docker Build
        run: |
            docker build --pull -t ${{ secrets.CONTAINER_REGISTRY_HOST }}/$ECR_REPOSITORY:latest .
            docker tag ${{ secrets.CONTAINER_REGISTRY_HOST }}/$ECR_REPOSITORY:latest ${{ secrets.CONTAINER_REGISTRY_HOST }}/$ECR_REPOSITORY:$ECR_IMAGE_VERSION
            docker tag ${{ secrets.CONTAINER_REGISTRY_HOST }}/$ECR_REPOSITORY:latest ${{ secrets.CONTAINER_REGISTRY_HOST }}/$ECR_REPOSITORY:beta

      - name: Push to ECR
        uses: olxbr/aws-ecr-push-action@v1
        id: ecr
        with:
          # The complete repository name from ECR {BU}/{TEAM}/{PROJECT} (ex. cross/devtools/devtools-scripts).
          ecr_repository: ${{ env.ECR_REPOSITORY }}
          # Comma-separated string of ECR image tags (ex. latest, 1.0.0)
          tags: 'latest,${{ env.ECR_IMAGE_VERSION }},beta'
          # Keep last N images on ECR (default: 100)
          keep_images: 20
        # Warning! Don't change this env values!
        # Just copy and paste this whole block as is in your repo
        env:
          AWS_ACCOUNT_ID: ${{ secrets.AWS_CROSS_ACCOUNT_ID }}
          AWS_ACCESS_KEY_ID: ${{ secrets.AWS_CROSS_ACCESS_KEY_ID }}
          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_CROSS_SECRET_ACCESS_KEY }}
          AWS_PRINCIPAL_RULES: ${{ secrets.AWS_PRINCIPAL_RULES }}

The image need to be at te format CONTAINER_REGISTRY_HOST/ECR_REPOSITORY:TAG For this, you can use the secret CONTAINER_REGISTRY_HOST to add this prefix.

Running local

You can run the action on your local machine with npm run start:local after exporting the correct AWS credentials.

Alternatively, you can use npm run stat:dry to start a dry run that will not push anything to AWS.